Dr. Dobb's Journal October 1998
As one who got sand kicked in his face at the beach and was always picked last when it came to choosing up sides for whateverkindaball, I acquired early on a vested interest in rooting for the little guy. I have watched both Revenge of the Nerds and the sequel, and I know the words to the Underdog theme song. I like Ally McBeal better than Murphy Brown.
I'm pleased when Microsoft at least temporarily fails to dominate a market niche or when it isn't allowed to buy the company that does. Although Jerry Sanders may be something of a sand-kicker himself, his AMD is an underdog and I'm glad he's kicking some sand back at beach bully Intel.
I truly believe that I live in a country blessed with the best, the most open government on Earth but I wouldn't turn my back on it, and when it comes to that government against one of its citizens, I'm with the citizen.
This month I'm just going to root for some little guys, underdogs, long shots, and Quixotic adventurers. Such as:
What is the hottest operating system today? Now, let's not always see the same hands.
BeOS, JavaOS, NT 5.0, no, those are fine answers, but not really what I was looking for. I guess I'd better tell you the answer I had in mind before someone says Windows 98 and I have to keep them after class to clean erasers.
My answer, of course, is DOS.
Zdnet Anchordesk editor Jesse Berst thinks it'll be the most talked-about operating system in 1999. Since 1997, he points out, Caldera has sold three million copies of DR-DOS, which it acquired from Digital Research byway of Novell.
As Jesse points out, Windows CE, JavaOS, and any other operating system that wants to power those set-top boxes, PDAs, smart phones, cash registers, kiosks, game machines, network computers and whatnot will have to be a better fit than DOS.
And Caldera has given DOS a graphical web browser that runs on minimal hardware with little RAM, multitasking, TCP/IP, HTTP, FTP, POP3 and SMTP, and power management. Now what if Caldera can cut licenses for network computers that will sell for less than Microsoft charges for a Windows CE license? Perhaps you thought I meant another underdog OS currently enjoying great popularity? Linux is hot, all right. Oracle, Informix, and Computer Associates all announced this summer that they would be porting major applications to Linux. Netscape has announced a Linux version of its software.
Meanwhile, the NT 5 release experiences delay after delay.
A feature in the latest release of Netscape Navigator raises some interesting questions, and created a problem for one developer who started out as a proponent of the feature.
Navigator has always allowed shortcutting in the specification of URLs. To get to http://www.swaine.com/, you could just type "www.swaine.com" or even just "swaine" and Netscape guessed what you had left out and took you where you wanted to go.
Usually. This shortcutting is a form of information reduction, and whenever you rely on reduced information it will eventually bite you.
Kids typing "whitehouse" at Navigator soon learned this, as they learned the difference between www.whitehouse.com, with its explicit pictures, and www.whitehouse.gov, which may be a source of salacious material but doesn't actually post nude GIFs.
Navigator 4.5 performs considerably more analysis on that shortcut you type in than the previous releases did, sifting through a series of databases to make a more educated guess about where you want to go today. It's a system that in theory gives preference to trademarks over domain names and domain names over subject areas in Netscape's site, although it didn't seem to work that way for one developer. Dave Winer of Userland Software, who owns the domain name scripting.com, was bitten by the new feature. Typing in "scripting" used to take him, and lots of other people, to his web site. When the Navigator 4.5 beta was released, Winer typed "scripting" and found himself at a Netscape site.
A lot of mostly polite discussion ensued, with some people questioning whether it was right for Winer to own a generic term like scripting. The essential point, though, it seems to me, is that Netscape did something you should never do -- it removed a popular feature from its software and put a new, similar, but confusingly different, feature in its place. I liked the ability to type in the essential element of a .com domain name to go there. I wouldn't mind also having a keyword lookup system. But I don't want to give up one for the other. Netscape should not break one to provide the other.
So far, this is a user-interface-design story. But there is a more disturbing angle to what Netscape did. The site to which Netscape linked Winer was one of Netscape's own. A feature purporting to be a neutral linking facility was in fact pumping up the portal's hit count, and hijacking visitors from others' sites. Or so one could argue. This example and one cited later in this column show that, on the Web, advertising often masquerades as editorial content or a navigational service.
What's on Winer's mind when he's not defending his domain name is XML. Specifically XML-RPC, a spec for remote procedure calls via XML, which Userland is working on with some other companies (read Microsoft). The arguments for a universal specification for a remote procedure call protocol are compelling: As Winer sees it, one such argument is empowering users to integrate products without having to wait for software companies to do it for them.
He cites an area where users have done just that. The integration of databases and web sites makes sublime sense, Winer says, "but all the pioneering work was done by users!" Database vendors didn't have any particular insight into the Web when it onto into the scene, but web developers had a compelling need to hook up to databases, and they figured out how to do it. It was easy because there are standard interfaces for databases, but with standard interfaces to program functionality, savvy users will be able to do more powerful integration. So goes the (familiar) argument. Speaking as a savvy user, I say, why not?
XML has some desirable features, but one compelling reason to support an XML-based RPC is that Microsoft seems committed to it. Microsoft has stated publicly that it will open up its operating system using XML-RPC. Example 1 is an XML-RPC request, courtesy of Winer's web site (http://www.scripting.com/, but don't just type "scripting").
An XML-RPC message is an HTTP-POST request. A procedure executes on the server and the value it returns is also formatted in XML. Procedure parameters can be scalars, numbers, strings, dates, and so on; and can also be complex record and list structures. As Example 2 shows, structs can be recursive.
<opinion>I see where we're headed. HTML was only the tip of the iceberg. In a couple of years we'll be as deep in angle brackets as any Lisp programmer ever was in parentheses.</opinion>
This seems to be the "Scripting Paradigms" section of this month's column. That being the case, here's another scripting item, and the only strictly Mac-oriented item in this column this month: SuperCard has a new home. Look for the scriptable, Mac-based, HyperCardlike multimedia development environment online at http://www.incwell.com/. That's the web site of IncWell, which is working on the 3.5 upgrade, aiming for delivery in September. The web site has the feature list and ordering instructions. IncWell is, by my count, the fourth owner of SuperCard, a very useful product that deserves to survive. Compared with the Amiga computer, another product that won't quite die, SuperCard looks positively radiant. The features of Version 3.5 make it look like anything but an old product that has been shunted from owner to owner for years. Somebody needs to finish the Window version, though.
I picked up the first issues of two new magazines recently. One, Business 2.0 from Imagine Media, is really The Net relaunched as a publication tracking the changes in how business is done as a result of the Internet. Definitely tangential to our interests here, it nevertheless does touch on some interesting topics, like an intro to XML and a brief look at Guy Kawasaki's Garage.com site.
Garage.com is Guy's attempt to bring together the garage startups with the money men who would otherwise never notice them. Venture capitalists like to get in on the second round of financing, after a company has had a chance to show what it will do with a little money. They aren't so eager to jump into absolute startups, so these companies often have to be touched by an angel -- maybe Aunt May, maybe Paul Allen -- for the initial seed money.
Guy's business tries to offer a service to each of its constituencies. For entrepreneurs, it offers advice, research, and connections with money people, the goal being to help them get that seed money. For investors, it screens and categorizes the startups and presents them as targeted investment opportunities.
It's just getting started. It seems clear that Guy has identified a need, or two complementary needs. If his company fills it well, it could be great for all concerned.
The other magazine I picked up was Brill's Content. It's a media watchdog magazine and the first issue looked promising. There were stories about good reporting and bad reporting, the latter exemplified by an endless, day-by-day breakdown of the Monica Lewinsky press coverage that reminded me of trying to cure someone of a nicotine addiction by rubbing his nose in an ash tray. Had enough yet?
More germane to our interests in the slice of our lives we address here in DDJ is the e-media coverage. Esther Dyson, a regular contributor, wrote about the futility of propaganda on the Internet, and there was a nice piece analyzing how much of the purported editorial content of the major portal sites is really paid positioning. It's not there because the site owner paid for it because it has value to you, it's there because the supplier paid the site owner to place it there to get the clicks. It's all about value to others, not to you.
It doesn't have to be that way. Let me tell you about a web site dedicated to reducing the commercial shackles of publishing, specifically book publishing.
Gregg Williams, who had a career at Byte magazine and at Apple and is now working on a third career, has noticed that the "Internet has the potential to...make every book that anyone wants to write available to anyone who wants to read it."
He's exactly right, and it could lead to, as he rightly points out, a revolution in publishing on a par with the invention of the printing press. Gregg has set up a web site, http://www.pubspace.com/, to explore the possibilities. No doubt amazon.com and barnesandnoble.com and heavyweights will be exploring the possibilities too, and reaching different conclusions, but as Gregg argues very convincingly, the Internet also enables the little guy.
Cyber Rights by Mike Godwin (Times Books/Random House, 1998, ISBN 0-8129-2834-2) is a book for any champion of the rights of the little guy. Godwin, counsel to the Electronic Frontier Foundation, set out in this book to explain cyberspace and why it matters; to show how the law and the U.S. Constitution apply to cyberspace, or how they should; to answer widespread fears about cyberspace; and to defend the Internet. I think he does a good job on all counts.
Godwin is not a dispassionate observer. If I read his biases correctly, he believes that it's better for children to live in an adult world than for adults to live in a child's world, better for the guilty to go free than for the innocent to be punished, that freedom is not a means to an end but a human need.
Godwin has been involved in some landmark cases involving the Internet, sometimes only peripherally. The book is organized around those cases, and although it does justice to the cases and to the law, it is hardly dry legalese. Godwin is a surprisingly readable writer. In part it's because he is passionate about what he's writing; in part, because the issues are so important. But it's also because he's a good writer. The book addresses the whole range of speech issues on the Net: libel, what it means, and whether the concept even applies to the Net; hate speech; privacy; copyright and other intellectual property issues; censorship, obscenity, and whether there's any sense in applying community standards to a global medium. The capstone of the book and of Godwin's career to date is the defeat of the Communications Decency Act of 1996, a triumph of common sense over FUD.
Mike (Swaine, not Godwin) maintains a support forum for this column, plus a daily tech news digest, at http://www .swaine.com/.
DDJ